&DRIFT: ARRIVAL-TIME-DISTRIBUTION
The electrons on which you trigger. These are usually small
numbers, around 5, but you are free to choose larger values.
You may list several electrons, up to a compilation dependent
limit (currently set to 10).
In the rest of the descriptions, these electrons are referred
to as the 'selected electrons'.
If a value is negative, then the electrons will be counted
from the back, the last electron corresponds to a value of 0.
For the 4 last electrons, you can also type LAST, ONE-BUT-LAST,
TWO-BUT-LAST and THREE-BUT-LAST.
Default: initially, only the 25th electron is selected. The
set of selected is remembered as default for the next call.
A probability between 0 and 1 (both excluded). Apart from
the mean arrival time of the selected electron, Garfield
outputs also the time at which the selected electron has a
given probability to have arrived.
Default: initially 0.5, the median of the distribution. The
value set is kept as default for the next call.
Whilst the program can compute reasonable bounds for the
overall arrival time distribution histogram, it can not do
so for the histogram of the selected electron.
Unless an explicit time window is specified, Garfield will
therefore take the computed time range for the all-electrons
arrival time histogram, and it will book an histogram with
automatic range calculation for the arrival time distribution
of the selected electron. Automatic range calculation means
that the first 100 entries are used to compute a suitable
range.
If you wish the selected electron arrival time distribution
histogram to have the same range as the histogram with the
distribution for all electrons, then you should specify the
NOAUTOSCALE-TIME-WINDOW option. Keep in mind, if you do so,
that the accuracy of the median arrival time of the selected
electron can be reduced because the arrival time distribution
is likely to cover only a small fraction of the range of the
histogram.
If you wish to specify a time window yourself, then use the
TIME-WINDOW keyword.
Default: automatic scaling is on by default.
Specifies the time window of all arrival time histograms, both
of the selected electrons and of all electrons.
This option is most useful if used in conjunction with the
KEEP-HISTOGRAMS option when you wish to study the histograms
in more detail.
Default: by default the time range of the selected electron
histogram is computed from the first 100 entries of this
histogram, whereas the range of the all-electron histogram
is set to the time range for the track under study.
The part of the AREA to be covered by the tracks.
Default: the entire x-range of the AREA.
The x-distance between 2 tracks.
Default: about 5 % of the horizontal scale.
The vertical extent of the tracks. This parameter is less
trivial than it might appear: if the range does not cover
the full acceptance boundaries of the wire, the program
will give wrong results for late electrons. If the range
is chosen too large, the waste of CPU time is considerable
and also the accuracy may suffer if the LINES parameter
does not compensate.
Default: the entire y-range of the AREA.
By default, diffusion is included in the arrival time
calculations. If you wish to study the impact diffusion
has on the distributions, then it may be convenient to
be able to switch diffusion off. This can be achieved
by selecting the NODIFFUSION option.
Depending on the presence of attachment data or not,
attachment is or is not included. If attachment data
is present, but if you do not want the effect to be
taken into account, then you can specify NOATTACHMENT.
Each track is first 'prepared': from regularly spaced points
on the track, electrons are drifted and the drift times and
integrated diffusion coefficients are stored for interpolation
when doing the Monte-Carlo cycles. Where the drift time
changes most between 2 points, additional drift lines are
calculated.
The parameter lines is the total amount of drift lines that
are calculated this way: 75 % in the first step and 25 % in
the second.
Default: the LINES parameter from the &DRIFT section.
The angle of the tracks with respect to the y-axis.
Default: initially 0 degrees. The angle that you set is
kept as default for the next call.
The number of bins in the histograms used to compute the
time at which the selected histogram has a given probability
to have arrived - the mean arrival time is computed directly
without histogram.
Default: initially half of MXLIST, usually 100-500 bins.
The value that you set is kept as default for the next
call.
This option can be used to store the arrival time histograms
for later study. The histograms are made accessible via global
variable names that are displayed while the command is carried
out (SEL_n, and ALL_n with n a counter).
A common application of this option seems to be the Gaussian
fit of the arrival time histograms. This can be done using the
FIT_GAUSSIAN procedure, or by a user program after writing the
histograms to disk with the WRITE_HISTOGRAM procedure.
If (1) the instruction is placed in the body of a DO loop and
(2) these global variables are used in arithmetic or in IF
conditions, then you have to declare the global variables before
the DO loop by assigning them an (arbitrary value). Declaring is
not required if the variables are only used outside a DO loop,
e.g. in SAY statements or in procedure calls.
Default: histograms are not kept. The setting is kept across calls
of ARRIVAL.
If this option is selected, then the x(t) relations that have been
obtained by ARRIVAL, will be stored as a set of matrices and numbers:
-----------------------------------------------------------------
Global Description Type
-----------------------------------------------------------------
X_wire x-coordinates Matrix
MEAN_wire mean time all electrons Matrix
MEDIAN_wire median time all electrons Matrix
RMS_wire RMS time all electrons Matrix
MEANelectron_wire mean time selected electron Matrix
MEDIANelectron_wire median time selected electron Matrix
RMSelectron_wire RMS time selected electron Matrix
E_electron sequence # selected electron Number
-----------------------------------------------------------------
The string "wire" in the global variable name is equal to the wire
number of the wire for which the x(t) relation is calculated.
The string "electron" is an index that runs from 1 to the number
of selected electrons. E_electron shows which electron a given
index corresponds to. This can be zero or a negative number: 0 means
the last electron, -1 the one but last etc.
If (1) the instruction is placed in the body of a DO loop and
(2) these global variables are used in arithmetic or in IF
conditions, then you have to declare the global variables before
the DO loop by assigning them an (arbitrary value). Declaring is
not required if the variables are only used outside a DO loop,
e.g. in SAY statements or in procedure calls.
This option is not on by default, but its setting is remembered
from one call to the next.
The number of Monte Carlo loops over track generation. These
loops consume relatively little CPU time since the drift times,
diffusion coefficients etc are all taken from interpolation
tables.
It seems therefore tempting to select a large value for this
parameter. When saving the distributions (KEEP-HISTOGRAM) and
then using them in a fit, this may indeed be a good approach.
However, when extracting information from the MC process with
means and RMS, the accuracy tends to degrade for large values
of loop, mainly because of single entries with very deviating
values.
When specifying SINGLE-CLUSTER, only one cluster is generated
per track, as a result the default value will be low and you
are advised to choose a larger value in this case.
Default: initially 1000. Each ARRIVAL statement has as default
the value set in the previous invocation.
The time by which the selected electron has a certain
probability to have arrived, is obtained by reverse polynomial
interpolation in the cumulative timing histogram. Using this
keyword, you can change the order of the polynomial.
Default: 2, parabolic interpolation. The setting will be
remembered from one call to the next.
The name of the dataset you wish to write your results to.
If you don't specify such a dataset name, you will only
receive plots and no numeric output.
Default: By default no dataset is written.
The name to be assigned to the member.
Default: none.
Garfield uses by default a string specifying the wire number
as remark for the member header. You may override this.
Garfield can be asked to output the following graphs for
each track:
-------------------------------------------------------------------
Option Plot produced
-------------------------------------------------------------------
PLOT-EACH-X-OVERALL The arrival time distribution of all
electrons from the track that crosses
a line parallel to the x-axis at the
value of x currently looked at.
PLOT-EACH-X-SELECTED-ELECTRON The arrival time distributions for
the selected electrons from the track
that crosses a line parallel to the
x-axis at the value of x currently
looked at.
-------------------------------------------------------------------
Both plots are by default not made.
In addition, plots can be requested that summarise the results of
the computation for each wire, as function of x. The PLOT-OVERVIEW
option will ensure that the following 3 graphs are made:
1. The mean drift time of all electrons and the mean drift time
of the selected electrons.
This graph is a more refined version of the x(t) relation as
calculated by the XT-PLOT instruction.
2. The drift time by which any electron has a probability of
'threshold' to have arrived and the drift time by which the
selected electrons have this probability to have arrived.
Since the 'threshold' is usually set to 50 %, this probability
is in practice equivalent to the median.
3. The spread of the arrival time distribution of all electrons,
and the spread of the arrival time distributions of the selected
electrons.
The spread in shown in this graph is the standard deviation of
the arrival time histogram. This measure is sensitive to
outliers, and should not be used for delicate studies of the
resolution. The KEEP-HISTOGRAMS option can be used to store
the histograms, more robust techniques can then be applied
outside Garfield.
In all plots, the all-electron graphs are plotted with the
graphics representation FUNCTION-1, while the graphs for the
selected electrons are plotted as FUNCTION-2.
These plots are initially made by default, but the default
setting of each call is the setting of the previous call.
The timing histograms for the selected electrons and for
all electrons are printed out at your request. In the printout,
additional information such as the mean and width are provided.
These quantities are computed from the entries themselves, not
from the accumulated bin contents. The internal arithmetic is
done in double precision.
The setting is remembered from one call to the next.
Keyword index.
Formatted on 10/11/98.